Wav2Letter: an End-to-End ConvNet-based Speech Recognition System
Preprint
Abstract
This paper presents a simple end-to-end model for speech recognition, combining a convolutional network based acoustic model and a graph decoding. It is trained to output letters, with transcribed speech, without the need for force alignment of phonemes. We introduce an automatic segmentation criterion for training from sequence annotation without alignment that is on par with CTC while being simpler. We show competitive results in word error rate on the Librispeech corpus with MFCC features, and promising results from raw waveform.Keywords
All Related Versions
This publication has 7 references indexed in Scilit:
- Very deep multilingual convolutional neural networks for LVCSR2016
- EESEN: End-to-end speech recognition using deep RNN models and WFST-based decoding2015
- Joint training of convolutional and non-convolutional neural networks2014
- Speaker adaptation of neural network acoustic models using i-vectors2013
- Estimating phoneme class conditional probabilities from raw speech signal using convolutional neural networks2013
- Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research GroupsIEEE Signal Processing Magazine, 2012
- Improvements in beam search1994